Health consequences of retirement have not been included in the current public debate about increasing the age at retirement, which might be due to the fact that studies aimed at health consequences of retirement show ambiguous results. The literature indicates that various contextual characteristics might explain conflicting results. The current study examines the effect of retirement and age at retirement (55-64 years) on self-perceived health. Characteristics tested for confounding and effect modification were: demographic, health, psychological, job, and retirement characteristics. Subjects were 506 participants in the Longitudinal Aging Study Amsterdam (LASA). After three years, 216 retired and 290 remained employed. Multinomial logistic regression analyses show no main effect for retirement compared to continued employment. Modal (59-60) retirees were more likely to attain excellent or good self-perceived health (less than good self-perceived health as the reference category). Early (55-58) and late (61-64) retirees were unaffected by retirement if they did not receive a disability pension. Early and late retirees who received a disability pension were less likely to attain excellent self-perceived health after retirement. Higher educated were less likely to attain excellent self-perceived health after retirement, especially at late retirement age, although health selection might explain this result. Finally, mastery possibly acts as an adjustment resource. The paper concludes with a discussion on explanations for the effect of retirement and age at retirement. © 2011 Cambridge University Press.