Loneliness has been associated with poor mental health and wellbeing. In England, a 2018 national strategy on loneliness was published, and public health guidelines recommend participation in social activities. In the absence of existing economic evidence, we modelled the potential cost effectiveness of a service that connects lonely older people to social activities against no-intervention. A 5-year Markov model was constructed from a health and social care perspective. Parameters were drawn from the literature, with the intervention structure based on an existing loneliness alleviation programme implemented in several settings across England. Univariate and probabilistic sensitivity analyses were undertaken. The total expected cost per participant in the intervention group is £ 7131 compared to £ 6783 in the usual care group with 0.45 loneliness free years (LFY) gained. The incremental cost per LFY gained is £ 768; in the probabilistic sensitivity analysis the intervention is cost saving in 3.5% of iterations. Potentially such interventions may be cost-effective but are unlikely to be cost-saving even allowing for sustained effects and cumulative adverse health and social care events averted. Empirical studies are needed to determine the cost-effectiveness of these interventions, ideally mapping changes in loneliness to the quality of life, in order to allow the key metric in health economic studies, cost per quality adjusted life year to be estimated.