Introduction: In the last decades, the areas of criminal recidivism and risk assessment have been gaining prominence in the scientific community 1 . The literature confirms we must assess the risk and protective factors to know how to intervene with those who are serving prison sentences 2 . It also points to the importance of certain variables such as age 3 and criminal versatility 4 . The main goals of this study are to characterize the risk and protective factors of a sample of prison inmates; to verify the association between the risk, current age and age at the time of the first arrest; and to analyse the differences in the risk according to criminal versatility. Materials and methods: The current study has an exploratory, quantitative and comparative design, with a non-probability convenience sample composed by 73 male inmates, who have more than one conviction leading to imprisonment. For data collection we used the Historical, Clinical and Risk Management Scheme 5 (HCR-20) and the Structured Assessment of Protective Factors for Violence Risk 6 (SAPROF), as well as a form to assess age and criminal versatility. The investigation was approved by the Ethics Committee of Fernando Pessoa University, as well as by both the General Management of Social Reintegration and Prison Services and Setúbal's prison establishment. Results: The results point to a high risk of recidivism among inmates (low risk - 15.1%; medium risk - 43.8% and high risk - 41.1%). There are some risk factors prominently present in the sample, such as previous violence, substance abuse, impulsiveness and negative attitudes. The absence of certain protective factors such as empathy, adaptative coping, self-control and motivation for treatment can also be noted. No association could be found between the inmates' current age and risk (r=−.18, p=.13), but we confirmed a negative association between age at first arrest and risk (r=−.34, p=.00). The were no statistically significant differences between the means of the criminal versatility groups for risk F(2.69) = 2.92, p=.06, but a negative association could be found between criminal versatility and risk (r=−.28, p=.02). Discussion and conclusions: Although most of the participants have a high or medium risk of recidivism, their risk factors and absent protective factors are mostly dynamic, leading us to the conclusion that intervention in these cases would lower the risk. The negative association between age at first arrest and risk, as well as the negative association between criminal versatility and risk are supported by the previous literature. Age and criminal versatility are two factors that deserve to be studied in future works. The present investigation raises some serious questions about the rehabilitative work that can be done in prison establishments.