The six-hour disruption to the social network's services on Tuesday was caused by a change of internal code that made the micro-blogging platform crash, Twitter has revealed.
The outage, the longest in the website's history, started at around 8.30am GMT with users unable to log into their accounts and access the website.
Twitter managed to fix the problem late in the afternoon saying a faulty software update was removed to make the site run smoothly again.
"The intermittent issue affecting some users between 00:40 to 06:50 PST (Pacific Standard Time) has now been resolved,” the company tweeted.
"The issue was related to an internal code change. We reverted the change, which fixed the issue. Thank you for your patience."
Users in Europe, the Middle East, Africa and North America reported problems with the computer-based platform as well as mobile-based apps. Twitter’s Tweetdeck app was also affected.
The glitch, which received its own #twitterdown hashtag, caused the company’s shares to drop by 7 per cent on the day of the outage. Wall Street has long worried about Twitter's stagnant growth in users and advertising revenue, and analysts said the outage added to the concern.
"The current market malaise and the recent site outages are compounding the negatives and having a very negative reaction on the shares," said Victor Anthony, Axiom Capital Management analyst.
After the initial outage experienced by users in Europe, Saudi Arabia and South Africa, the site went partially back online, but disruptions continued throughout the day and spread to North America as well.
At approximately 17.45 GMT Twitter reported that some users were still having trouble accessing the service.
15 minutes later the company announced the service problems had been resolved.
Twitter currently has just over 300 million users, but had its slowest user growth in 2015. It was eclipsed by photo-sharing app Instagram, which is owned by Facebook and surpassed 400 million users last year.