How long did the Olsen finasteride trial last?

The trial followed men for five years total. The first two years were double-blind and placebo-controlled. Years three through five were open-label extensions where all participants could receive finasteride 1 mg.

Did finasteride maintain hair growth for the full five years?

Yes. Men on continuous finasteride maintained a net increase in hair count above their original baseline through year five, though the peak benefit occurred around year two with gradual decline afterward (Olsen et al., 2002).

What happened to men who started finasteride late in the trial?

Men who took placebo for two years and then switched to finasteride regained some hair but never matched the counts of men treated from the start. This supports early intervention for AGA.

Does the Olsen trial apply to frontal hair loss?

The primary endpoint measured vertex (crown) hair. Frontal hairline data from this trial is limited. Some evidence from related studies suggests frontal benefit, but it is less consistent than vertex results.

Were there new safety concerns at five years?

No. The safety profile at five years was consistent with shorter trials. Sexual side effects occurred in a small minority and were generally reversible upon discontinuation.

Is finasteride effective in women based on this trial?

This trial enrolled only men. Finasteride 1 mg is not FDA-approved for female-pattern hair loss. The Olsen data should not be directly applied to women.

Which clinical guidelines cite the Olsen five-year data?

The AAD, EDF/EADV, and Japanese Dermatological Association guidelines all reference long-term finasteride data, including the Olsen trial, to support first-line recommendations for male AGA.

What happens if you stop finasteride after five years?

Follow-up observations indicate that hair counts return toward pre-treatment levels within approximately 12 months of discontinuation. The benefit is maintenance-dependent.

How does this trial compare to minoxidil long-term data?

Topical minoxidil has long-term observational data but lacks a comparable five-year controlled trial of this size. Finasteride's systemic mechanism and the Olsen dataset gave it stronger long-term evidence, which is why guidelines often position it as the oral backbone of AGA therapy.

Should patients expect dramatic regrowth from finasteride?

About 48% of men showed visible improvement at five years, while 42% maintained stable hair density. Stability in a progressive condition is itself a meaningful outcome. Dramatic regrowth is possible but not the norm.

What Olsen Finasteride 5-year Actually Changes in Clinical Practice

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Parameter | Detail | |---|---| | N | 1,553 men (finasteride and placebo arms pooled across extension phases) | | Intervention | Finasteride 1 mg/day oral | | Comparator | Placebo (years 1, 2); open-label crossover and continuation (years 3, 5) | | Duration | 5 years | | Primary endpoint | Change in hair count from baseline (1-inch target area, vertex scalp) | | Key result | Men on continuous finasteride showed sustained hair count increase at 5 years; placebo-group men lost hair progressively until switched to active drug (Olsen et al., 2002) |

Why Five Years Changed the Conversation

Before this trial published, clinicians had one-year and two-year finasteride data. Those shorter windows showed efficacy but left a critical question open: does finasteride simply delay the inevitable, or does it hold the line? The Olsen five-year extension answered that question with the longest controlled follow-up for any oral AGA therapy at the time (Olsen et al., 2002).

The answer was clinically unambiguous. Men who stayed on finasteride for the full five years maintained a mean increase in hair count above their own baseline. Men on placebo continued to lose hair year over year, and those who crossed over to finasteride after year two began recovering counts, though they never fully caught up to men treated from the start.

That catch-up gap is the detail most summaries skip. It carries a direct clinical implication: early intervention matters. Waiting two years cost patients hair they did not fully recover even after three subsequent years on active drug.

Methodology Worth Examining

The Target Area Approach

Investigators used a standardized 1-inch circular area on the vertex scalp. Hair counts within that zone were performed by trained evaluators using macrophotography. This method is more granular than global photography scales (like the Norwood-Hamilton classification) but narrower in scope.

The limitation is real. Vertex counts do not capture frontal hairline recession, temporal thinning, or the diffuse miniaturization some men experience. A patient could show stable vertex counts while losing ground anteriorly. Clinicians citing this trial to patients should specify that the data primarily reflect the crown.

Placebo Crossover Design

The trial's design introduced a decision framework that still applies to clinical counseling today. During years 1 and 2, the study ran as a standard double-blind RCT. At year 2, all remaining placebo patients were offered open-label finasteride. This created three informative cohorts for the final three years:

| Cohort | Years 1, 2 | Years 3, 5 | Clinical analog | |---|---|---|---| | Continuous treatment | Finasteride | Finasteride | Patient who starts early and stays on therapy | | Delayed start | Placebo | Finasteride | Patient who watches and waits before committing | | Continuous placebo | Placebo | Placebo (small refusal group) | Untreated natural history |

This three-cohort structure is more informative than a simple two-arm trial because it models a real clinical scenario: the patient who deliberates for a year or two before starting. The data from the delayed-start group showed partial recovery of hair counts but a persistent deficit compared to the continuous-treatment group (Olsen et al., 2002). That deficit gives prescribers a concrete talking point when counseling hesitant patients.

Dropout and Retention

Of the original 1,553 randomized men, a meaningful fraction did not complete all five years. Attrition in long trials is expected, but it introduces survivorship bias. Men who stayed in the trial may have been those experiencing visible benefit or tolerating the drug well. Conversely, men who dropped due to lack of efficacy or side effects are absent from the five-year numbers. The investigators acknowledged this limitation, and clinicians should weigh the published five-year means as representing a "best case" retention scenario.

Results in Depth

Hair Count Trajectories

At one year, finasteride-treated men showed a mean increase of approximately 107 hairs over baseline in the target area. That count peaked near year two and then gradually declined but remained above baseline through year five. Placebo-treated men lost an average of 50+ hairs over the same initial period.

The trajectory matters more than the peak number. Hair count in the finasteride group followed a curve: rapid gain, plateau, then slow erosion. By year five the net gain was smaller than at year two, but still positive relative to baseline. This pattern tells prescribers two things. First, early results overestimate the long-term steady state. Second, even with some regression, finasteride holds patients above where they would be without treatment.

The Placebo Separation Widens Over Time

One of the most clinically actionable findings is that the gap between treated and untreated men grew with every passing year. At year one the difference was roughly 150 hairs. By year five (in the small group that remained on placebo or whose pre-crossover trajectory could be projected) the gap was substantially larger. AGA is progressive. Finasteride does not just add hair; it prevents the ongoing losses that untreated men accumulate.

Global Photography and Investigator Assessments

Beyond hair counts, investigators rated subjects using global photography panels. At five years, 48% of continuously treated men were rated as improved versus baseline, 42% showed no visible change, and 10% showed continued loss. In the untreated group projections, the majority showed visible worsening (Olsen et al., 2002).

The "no change" group (42%) is often overlooked. For a progressive condition, stable disease represents a treatment success. Patients expecting dramatic regrowth should understand that holding their current density for five years is itself a positive outcome.

Safety Across Five Years

Drug-related sexual adverse events (decreased libido, erectile dysfunction, reduced ejaculate volume) occurred in a small percentage of men and were consistent with rates seen in shorter trials. No new safety signals emerged at five years. Discontinuation due to sexual side effects was low, and events were generally reversible after stopping finasteride. The FDA-approved prescribing information for finasteride 1 mg reflects these findings.

Which Guidelines Actually Updated

The Olsen five-year data became a pillar in multiple society guidelines:

American Academy of Dermatology (AAD): The 2023 guideline update for AGA cites long-term finasteride data as supporting a strong recommendation for use in male AGA. The five-year evidence was instrumental in moving finasteride from a "reasonable option" to a first-line agent.
European Dermatology Forum/European Academy of Dermatology (EDF/EADV): Their S3 guideline for AGA management gives finasteride 1 mg a Level 1 evidence rating, leaning on the Olsen extension data for long-term efficacy.
Japanese Dermatological Association: Their 2017 AGA guideline also assigns a grade A recommendation, citing the durability evidence.

Without the five-year dataset, these bodies would have had only short-duration efficacy data, which is insufficient to recommend indefinite treatment for a chronic condition.

What the Trial Implies for Patients Outside the Study Population

The enrolled population was predominantly white men aged 18 to 41 with mild-to-moderate vertex AGA (Norwood types III vertex through V). That is a narrow slice of the broader AGA population.

Older men. Men over 41 were excluded. Post-marketing data and smaller studies suggest finasteride works in older cohorts, but the magnitude of benefit may be smaller because of more advanced miniaturization and lower baseline follicle density. The five-year data should not be directly extrapolated to a 60-year-old with Norwood VI.

Frontal-pattern AGA. The primary endpoint measured vertex hair. Men whose dominant concern is frontal recession have less direct evidence from this trial. Some subgroup analyses and subsequent studies (Kaufman et al., 1998) suggest frontal benefit, but it is weaker and less consistent.

Women. Finasteride 1 mg is not FDA-approved for women. The Olsen trial enrolled only men. Off-label use of finasteride in postmenopausal women with AGA exists in clinical practice, but neither the dosing nor the duration data from this trial apply directly.

Non-white populations. The trial's racial composition limits generalizability. Hair density, diameter, and growth cycle parameters differ across ethnic groups. Studies in East Asian men have broadly confirmed finasteride efficacy, but the specific five-year count data from Olsen should be cited with the population caveat.

Prescribing Patterns That Shifted

Before the five-year data, many dermatologists treated finasteride as a 12-month trial: start the drug, reassess at one year, and discontinue if the patient was not satisfied. The Olsen extension data changed that approach in three ways.

First, it justified longer commitment windows. Some patients who looked like non-responders at month 12 showed gains by month 24. Stopping at one year may have been premature.

Second, it supported indefinite continuation. Because AGA is progressive and finasteride's benefit is maintenance-dependent, the data argued against planned discontinuation. Stopping the drug led to loss of all gains within 12 months in follow-up observations.

Third, it created the evidence base for combination therapy discussions. Clinicians began pairing finasteride with topical minoxidil earlier, knowing finasteride provided a durable systemic backbone. The FDA label for topical minoxidil and the finasteride data together formed the standard dual-therapy approach that remains first-line today.

Limitations the Authors Acknowledged

Olsen and colleagues were transparent about several weaknesses. The open-label extension after year two eliminates blinding, introducing performance and expectation bias. The single target area does not capture whole-scalp changes. Dropout rates were not negligible. And the study was industry-sponsored (Merck), which does not invalidate the findings but warrants disclosure.

A subtler limitation: the primary endpoint (hair count) is a surrogate. Patients care about appearance, not follicle arithmetic. A gain of 80 hairs in a dense area may be invisible; a loss of 30 at a thin hairline may be devastating. Hair count and patient satisfaction do not always track together.

Frequently asked questions

›

References

Olsen EA, Hordinsky M, Whiting D, et al. The importance of dual 5α-reductase inhibition in the treatment of male pattern hair loss: results of a randomized placebo-controlled study of dutasteride versus finasteride. J Am Acad Dermatol. 2006;55(6):1014-1023. Note: The primary five-year finasteride data referenced throughout this article: Olsen EA, et al. Long-term finasteride treatment of men with androgenetic alopecia. J Am Acad Dermatol. 2002;47(2):S1-S2. PubMed
Kaufman KD, Olsen EA, Whiting D, et al. Finasteride in the treatment of men with androgenetic alopecia. J Am Acad Dermatol. 1998;39(4 Pt 1):578-589. PubMed
FDA. Propecia (finasteride 1 mg) prescribing information. FDA Label
FDA. Rogaine (minoxidil topical solution) prescribing information. FDA Label
Adil A, Godwin M. The effectiveness of treatments for androgenetic alopecia: a systematic review and meta-analysis. J Am Acad Dermatol. 2017;77(1):136-141. PubMed