Background People with disabilities experience higher rates of social isolation and loneliness than people without disabilities, but there is limited information about how these conditions are manifested in people with specific types of disabilities. Objectives Using data collected as part of the second administration of the National Survey on Health and Disability (n = 2,132), our objectives were to determine if disability type and recruitment method added explanatory power to observed levels of social connectedness, after controlling for socio-demographic and environmental indicators. Methods We used hierarchical regression to evaluate how socio-demographic, disability, environmental, and recruitment type explained four outcome variables for different dimensions of social connectedness, including satisfaction with social activity, quantity of social connections, quality of social connections, and loneliness. Results Across all models, employment status (not employed), having mental illness/psychiatric disability, lack of transportation for social needs, and recruitment type (MTurk) significantly predicted lower levels of social connectedness. Conclusions The study provides evidence that using alternative methods, such as MTurk, to complement conventional recruitment strategies can improve understanding of social isolation and feelings of loneliness among people with disabilities, particularly those with mental illnesses and/or psychiatric disorders who are less likely to be accessed through group affiliations, but who are at greater risk for experiencing social isolation and feelings of loneliness.