Spinal ultrasonography provides guidance for epidural insertion in obstetric patients. The primary objective of the study was to develop a training program in spinal ultrasound for anaesthetists and to determine its effect on the skill acquisition of anaesthetists with no prior spinal ultrasound experience. Eighteen anaesthetists underwent two structured workshops (one week apart), each followed by a practice session and videorecorded assessments. Participants were randomised to a protocol-driven or non-protocol driven spinal ultrasound teaching program. Two experts rated each individual's performance using a global rating scale (GRS), checklist and image quality scale. The primary outcome was the mean difference in GRS score between the two workshops, analysed using linear mixed models. Intraclass correlation coefficients were calculated to assess agreement between assessors' ratings. A total of 108 ultrasound scans were performed on five pregnant volunteers during the assessment periods. After adjusting for confounders, GRS scores increased on all three rating scales at the second workshop, this increase being 6.01 points (95% confidence interval 4.56 to 7.46, P<0.001) from a mean score of 28.4 (95% confidence interval 24.8 to 32.0). There was no significant difference in the scores between the two teaching groups (difference in GRS scores=1.36 points, 95% confidence interval -0.77 to 3.50, P=0.211). Intraclass correlation coefficients showed substantial assessor agreement for all three assessment methods (range 0.59 to 0.89). The results demonstrate that programmed spinal ultrasound training sessions involving practice with guidance and feedback from an expert, whether protocol-based or non-protocol based, lead to improved performance.