Orderly onset of psychiatric symptoms has implications for both case detection and the construct validity of the underlying illness. Mean age and survival-hazard techniques were used to study the onset of alcohol abuse and dependence (as defined in the 4th edition of the Diagnostic and Statistical Manual of Mental Disorders; American Psychiatric Association, 1994) in 369 clinical cases drawn from a heterogeneous regional sample. The methods provided a similar general pattern of symptom sequencing, though only survival-hazard analysis described a punctuated onset of alcoholism in 3 discrete stages: alcohol abuse, dependence, and accommodation to the illness. This model survived a rigorous program of tests for goodness of fit and described the majority of the sample, supporting the construct validity of both alcohol abuse as a discrete first illness phase and of dependence as a set of core constructs distinct from and succeeding abuse. The specific strengths of survival-hazard analysis as a research tool in illness staging research are discussed.