Using nationally representative data from Algeria and Jordan, this paper shows that social networks are crucial in labour market intermediation in Arab countries. We make use of binary and ordered probit regressions, corrected for sample selection using the Heckman model, to investigate determinants of job search intensity and determinants of the probability of finding a job through social contacts. After factoring the sample selection, our findings suggest that the use of population density as a proxy for the size and strength of social networks may only be appropriate for the studies of minorities and immigrants. We propose that strong ties (closer friends and relatives, and maybe friends on social media) may be more crucial in job finding than weak ones (number of inhabitants in adjacent areas). On average, the analysis shows that job search is more intensive in Jordan compared to Algeria. Among others, household wealth, the local unemployment rate, region, previous labour market experience, and to some extent education, appear to exert significant roles in determining intensity. Importantly, the study finds that social networks are a popular method of finding a job in Algeria and Jordan, but not for skilled jobs. Such methods increase the probability of obtaining less secured informal jobs. Finally, the study also shows that despite the importance of public sector agencies in the job search process, less than 5% in Algeria and 9% in Jordan of the young employed state that such agencies have helped them transit into employment.