OBJECTIVES: ADHD medication users frequently take ‘drug holidays’ during the summer months.  The study objective was to compare ADHD medication adherence for an entire year (EY-365 days) and the school year (SY-270 days) by medication class and type. METHODS: Continuously enrolled Texas Medicaid children (≤18 years) who had ≥2 prescription claims for an ADHD medication served as the study population.  EY and SY (September 1-May 31) prescription claims were extracted from July 2002-December 2008.  Prescription claims were grouped by medication class [immediate release(IR), extended release(ER), long-acting(LA), non-stimulant(NS)]; and medication type [stimulant(S) vs. non-stimulant(NS)].  Adherence, assessed by medication possession ratio (MPR) using a fixed interval denominator, was measured both continuously and dichotomously (80%).  T-tests, ANOVAs and chi-square were employed to determine differences between groups. RESULTS:   Overall mean adherence for EY (n=62,789) was 49.8±30.3 and SY (n=50,842) was 62.2±26.2.  For EY, mean medication class adherence was not significantly different between NS (52.5±30.9) and ER (52.1±30.2); however, LA (47.6±30.9) and IR (37.2±27.1) were significantly lower (p<0.0001).  Regarding SY, adherence was not significantly different among ER (63.7±26.0), LA (63.5±23.8) and NS (62.9±27.0), while IR (52.8±24.7) was significantly lower (p<0.0001).   When adherence was dichotomized, EY medication class adherence differed significantly (p<0.0001): NS(25.8%), ER(24.1%), LA(21.2%), IM(9.8%).  Similarly, SY differed significantly (p<0.0001): NS(30.8%), ER(30.0%), LA(25.6%), IM(16.2%).  NS had significantly higher mean adherence than S, respectively: EY(52.5±30.9 vs. 49.4±30.2; p<.0001) and SY(62.9±27.0 vs. 62.1±26.1; p<.01).  When dichotomized, results were similar (p<.0001): EY(25.8% vs. 21.5%) and SY(30.8% vs. 27.9%). CONCLUSIONS:   Subjects were more adherent during SY compared to EY. Medication class mean adherence differed depending on timeframe used.  For analyses comparing NS and S, NS had significantly higher adherence, however for SY mean adherence, the difference may not be practically significant.  Due to unique patient medication-taking behaviors, ADHD medication adherence differs depending on the timeframe used.