To explore how age and length of abstinence affects neuropsychological (NP) recovery in chronic male alcoholics, we examined 97 recently detoxified alcoholics (RDA) who were abstinent a mean of 29.7 (SD = 9.4) days at their first evaluation, and again 2 years later, at which time 62 had resumed drinking (RES) and 35 had maintained interim abstinence (ITA). To separate practice effects from NP recovery, the NP performance of the RES and ITA groups were compared to age- and education-matched 29 long-term abstinent (LTA) alcoholics, abstinent a mean of 4.3 (SD = 3.5) years at initial testing, and 49 nonalcoholic controls (NAC). Sample mean age and education was 48.7 (SD = 8.8) and 13.9 (SD = 2.7) years, respectively. The RDA and LTA groups were also matched on years of alcoholic drinking [i.e., 17.2 (SD = 9.3) and 15.6 (SD = 8.7) years, respectively]. Results of 14 NP tests were reduced to 6 NP abilities (i.e., attention, learning, memory (recall), abstraction–cognitive flexibility, complex perceptual–motor integration and simple motor skills) that formed dependent variables in 4 groups × 2 ages (median split at 51.4 years classified participants as young and old) repeated measures MANOVA. Results: (1) the ITA group improved on abstracting ability more than other groups (p = .009); (2) the RES group deteriorated on motor tests (p = .03); (3) all younger alcoholics exceeded practice effects on abstracting ability, while older RES participants deteriorated relative to ITA group (3-way interaction, p = .03); and (4) the LTA group was comparable in NP performance to the NAC group. In conclusion, the extent and features of NP recovery among alcoholics are related to age, length of abstinence, and interim drinking. (JINS, 1999, 5, 234–246.)