Objective Assess the feasibility of recruiting children with special health care needs (CSHCN) into a clinical study using Medicaid data. Methods We used the 3M Clinical Risk Grouping software to identify and classify Medicaid-enrolled CSHCN ages 7-17 years in Washington into one of three chronic condition severity groups: episodic, life-long, or malignancy/catastrophic. We compared contact, recruitment, enrollment, and data completion proportions across severity groups using the Pearson chi-square or Fisher’s exact test. Results Of the 2,666 CSHSN in the Medicaid files, we contacted 922 and recruited 211 CSHCN (22.8 percent). One-hundred-thirteen made their study visit (53.6 percent) and 98 had complete data (86.7 percent). We contacted and recruited significantly larger proportions of CSHCN with malignancies/catastrophic conditions (P < 0.01). Across severity groups, there were no differences in enrollment (P = 0.27), but a smaller proportion of CSHCN with life-long conditions had complete data (P = 0.03). Conclusion It is feasible to recruit CSHCN from Medicaid files for clinical studies.