BACKGROUND: Social cognitive variables are often examined for their association with initial participation in colorectal cancer screening. Few studies have examined the association of these variables with adherence to multiple screening offers i.e., rescreening. This study aimed to describe patterns of participatory behaviour after three rounds of screening using faecal immunochemical tests (FIT) and to determine social cognitive, demographic and background variables predictive of variations in adherence. METHODS: Participants were 1,540 men and women aged 50 to 75 living in South Australia who completed a behavioural survey measuring demographic (for example, age, gender) and social cognitive variables relevant to FIT screening (for example, perceived barriers, benefits, self-efficacy). The survey was followed by three, free FIT screening offers mailed on an annual basis from 2008 to 2010. Patterns of participation after three screening rounds were described as one of five screening behaviours; 1) consistent re-participation (adherent with all screening rounds), 2) consistent refusal (adherent with no screening rounds), 3) drop out (adherent with earlier but not later rounds), 4) intermittent re-participation (adherent with alternate rounds) and 5) delayed entry (adherent with later but not initial round(s)). Univariate (Chi Square and Analysis of Variance) and multivariate (Generalised Estimating Equations) analyses were conducted to determine variables predictive of each category of non-adherence (those that did not participate in every screening offer, groups 2, 3, 4 and 5) relative to consistent re-participation. RESULTS: Significant social cognitive predictors of non-adherence were; less self-efficacy (drop out and consistent refusal), greater perceived barriers (drop out) and lower levels of response efficacy (consistent refusal). Demographic predictors of non-adherence included; male gender (delayed entry), younger age (intermittent, delayed and consistent refusal), less frequent GP visits (intermittent re-participation) and 'ancillary only' private health insurance (drop out). Less satisfaction with screening at baseline predicted drop out, consistent refusal and delayed entry. CONCLUSIONS: Different combinations of demographic and behavioural variables predicted different patterns of rescreening adherence. Rescreening interventions may benefit from a targeted approach that considers the different needs of the population subgroups. Satisfaction with past FOBT screening measured prior to the study screening offers was an important predictor of adherence.