Self-report retrospective estimates of sleep behaviors are not as accurate as prospective estimates from sleep diaries, but are more practical for epidemiological studies. Therefore, it is important to evaluate the validity of retrospective measures and improve upon them. The current study compared sleep diaries to two self-report retrospective measures of sleep, the commonly used Pittsburgh Sleep Quality Index (PSQI) and a newly developed sleep questionnaire (SQ), which assessed weekday and weekend sleep separately. It was hypothesized that the new measure would be more accurate than the PSQI because it accounts for variability in sleep throughout the week. The relative accuracy of the PSQI and SQ in obtaining estimates of total sleep time (TST), sleep efficiency (SE), and sleep onset latency (SOL) was examined by comparing their mean differences from, and correlations with, estimates obtained by the sleep diaries. Correlations of the PSQI and SQ with the sleep diaries were moderate, with the SQ having significantly stronger correlations on the parameters of TST, SE, and sleep quality ratings. The SQ also had significantly smaller mean differences from sleep diaries on SOL and SE. The overall pattern of results indicated that the SQ performs better than the PSQI when compared to sleep diaries.