We investigated efficacy and toxicity of replacing conventional triple (cytarabine, methotrexate, and hydrocortisone) intrathecal therapy (TIT) with liposomal cytarabine during maintenance therapy among 40 acute lymphoblastic leukemia patients. Twenty-eight of 29 patients in the TIT arm received TIT and 9/11 in the liposomal cytarabine arm received liposomal cytarabine. Arachnoiditis occurred in all initial 5 patients given liposomal cytarabine and intrathecal prednisolone succinate. Subsequently liposomal cytarabine was given with systemic dexamethasone. Neurotoxicity occurred at 6/27 liposomal cytarabine administrations with concomitant dexamethasone (22%). More liposomal cytarabine-treated patients experienced neurotoxicity in relation to intrathecal therapy during at least 1 cycle compared with TIT-treated patients (6/9 [67%] vs. 3/28 [11%], P=0.002). Apart from intermittent lower extremity sensory pain in 1 liposomal cytarabine-treated patient, no permanent adverse neurological sequelae were observed. In intention-to-treat analysis, projected 5-year event-free survival (pEFS-5y) was borderline higher for patients in the liposomal cytarabine arm compared with the TIT arm (1.0 vs. 0.69, P=0.046). However, pEFS-5y and projected 5-year relapse-free survival did not differ signficantly between patients treated with liposomal cytarabine or TIT (1.0 vs. 0.73, P=0.10; 1.0 vs. 0.76, P=0.12). Larger prospective trials are needed to explore whether liposomal cytarabine should be used as first-line prevention of relapse.