Background: Chronic medical illnesses coexist with mental health challenges, negatively impacting quality of life and well-being. Compassion-based interventions have shown promise for individuals with chronic conditions, yet accessibility barriers limit their implementation. Internet-delivered formats may address these limitations while maintaining effectiveness. To our knowledge, no fully self-guided, internet-delivered attachment-based compassion intervention has been tested in a transdiagnostic chronic illness population. Objective: This study aimed to evaluate the efficacy of internet attachment-based compassion therapy (iABCT) in improving quality of life and well-being (primary outcomes), and secondary psychological variables in adults with chronic medical conditions, compared to a waiting list (WL) control, and to assess acceptability and implementation barriers. Methods: A 2-arm, parallel-group, exploratory randomized controlled trial was conducted in Spain, with open-access recruitment through online and offline channels, and assessments completed via web-based self-report questionnaires. Adults (≥18 years) with a chronic medical condition and ability to read Spanish were eligible; individuals with terminal illness, current psychotherapy, or severe mental disorders were excluded. Participants were randomly allocated (1:1) to iABCT or a WL control (no intervention for 3 months) using a stratified, centralized computer-generated sequence. iABCT is a fully self-guided, 8-module online program delivered over 8 weeks, grounded in attachment theory and compassion-focused principles. Blinding of participants was not possible due to the nature of the intervention. Primary outcomes were quality of life (EQ-5D) and well-being (Pemberton Happiness Index [PHI]), assessed at baseline, 3 months, and 6 months. Data were analyzed using linear mixed-effects models under an intention-to-treat approach. Results: A total of 146 participants were randomized (iABCT n=72; WL n=74), all included in intention-to-treat analyses, with data available for 27 iABCT and 26 WL participants at 3-month follow-up. Between-group comparisons favored iABCT for overall quality of life (Cohen d=0.62, 95% CI 0.07-1.17) and preoccupied attachment, while well-being and most secondary outcomes showed within-group improvements without significant between-group differences. Quality of life improvements were maintained at 6 months within the iABCT group. No serious adverse events were reported in either group. Participants reported high satisfaction and usability. Qualitative interviews identified lack of therapist contact and insufficient monitoring as main engagement barriers. Conclusions: This trial provides preliminary evidence that iABCT may improve quality of life in adults with chronic medical conditions. It represents the first fully self-guided, internet-delivered attachment-based compassion intervention tested across diverse chronic conditions, uniquely integrating attachment security as a theoretical change mechanism alongside self-compassion-an approach that may address access barriers related to geography, mobility, or socioeconomic constraints. However, findings remain exploratory given high attrition, modest sample size, and no active control. Future adequately powered trials incorporating minimal human support, active control conditions, and cost-effectiveness analyses are needed to optimize real-world implementation.